The health and long-term survival of the growing population of childhood cancer survivors is threatened by late sequelae of therapy (cardiovascular disease, obesity, osteoporosis, new malignancies); health risking behaviors can significantly increase survivors' vulnerability to these late effects, while selected health protective behaviors may significantly reduce their risk. The Institute of Medicine has recently challenged investigators to develop interventions to ameliorate the late effects of therapy for childhood cancer survivors. However, there is little or no information about the antecedents of health risking and health protective behavior in this population, and the interventions designed to modify survivors' behaviors have produced disappointing results. The proposed study is a secondary analysis of the U.S. and Canadian-based Childhood Cancer Survivors Study (CCSS) aimed at generating new hypotheses about the antecedents of health-risking behavior and health -protective behavior in the largest and most diverse sample currently available to investigators. Specifically, the project proposes to: (1) Document the individual, cumulative, and interactive effects of selected demographic, diagnosis/treatment-related, resource, social, cognitive, affective, and motivational variables on 3 specific health-risk behaviors (alcohol use, smoking, and other tobacco use) and on 9 specific health-protective behaviors (physical activity, sunscreen use, breast/testicular self exam, breast/testicular provider exam, mammogram, Pap smear, medical check-ups); (2) Identify health-risk and health-protective behaviors that occur in related clusters, and determine how these clusters are related to age, gender, and social context sub-groups; and (3) Construct separate structural equation models (health-risk behavior and health-protective behavior) for survivors of childhood cancer by testing structural hypotheses derived from the Interaction Model of Client Health Behavior (IMCHB). Support for this study will afford an established investigator an opportunity to enter a newly targeted high-priority area of interest with a self-contained research project. By using a broad-based theory of health behavior and advanced analytic techniques, this study has the potential to make a major conceptual contribution to childhood cancer research and to inform the conceptual foundation and methodology of prospective clinical trials of health behavior interventions.